CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Microchannels Reactor (

JoanaAndrade March 4, 2013 13:39

Microchannels Reactor
I'm new in CFX and I have to model a microchannel reactor. It has an cylindrical inlet (7,2 mm diameter) connected with a cubic part (10 mm side). From this cubic part to the outlet (that is symmetrical) there are a lot of microchannels (1mm side). After setting up, when I'm gonna see the results, there is no way the flow goes through the microchannels.
Can someone help me?

ghorrocks March 4, 2013 17:26

Please post an image of what you are modelling, and some details about the fluid and boundary conditions.

JoanaAndrade March 6, 2013 13:55

1 Attachment(s)
It is like that. First, I'm trying to make a water flow. In the inlet, the flow rate range is between 0.25 and 250 g/min. In the outlet, relative pressure = 0.

ghorrocks March 6, 2013 18:15

Is it correct that one fluid flows through the orange channels and another through the blue channels and they cannot mix? If so then this is a trivial flow to model. There will be something simple wrong with your set up. If it is more complex then this please explain in more detail.

JoanaAndrade March 7, 2013 15:05

1 Attachment(s)
Yes, they cannot mix. I know it seems trivial, but the fluids just don't go through the channels. It goes like that. The set up is very simple, It's just water and air, no heating. And I've already increased the flow rate, and it was just the same. I'm wondering if there something wrong with meshing. I've just generate the standard mesh, no changes.

evcelica March 7, 2013 17:20

Looks like it goes through the channels to me? Inlets are on the top and the left? Maybe you think it doesn't because the streamline don't go all the way through? That doesn't mean the flow stopped.

JoanaAndrade March 11, 2013 11:55

Really? I tought so... And how do I know it is ok? Isn't the streamlines supposed to show the way the flow travels?

oj.bulmer March 11, 2013 16:09

If I am not missing something, this seems perfectly normal. If you have fully converged the solution, what is the domain imbalance for both the fluids? If it is less than (at least) 1% it should be alright. For better results, I would run till the domain imbalances are less than 0.01%, if it can be afforded.


ghorrocks March 11, 2013 17:38

Yes, the streamlines show the flow travel. But they are numerically integrated over the solution dataset and therefore are approximations. When you have long thin channels they do tend to wander a little and if they hit a wall they will stop.

All times are GMT -4. The time now is 19:32.